Chris Game wrote:

> Karl von Laudermann wrote:
> 
> 
>>Heh. I already wrote a Sudoku solver back in May; all I have to do
>>is change the input/output format a bit to match the quiz
>>example. In the meantime, here's a harder puzzle to test your
>>programs with:
> 
> 
> Why is it harder? If an algorithm works, all solvable puzzles are
> the same?

This is true for brute force solving only. If you apply some logic
to cut down the calculation time things gets more interesting.

> There's a soln on RubyForge somewhere I saw the other day...

Maybe, but its fun.

To those who try, here is a very neat one:

+-------+-------+-------+
| _ _ _ | _ 7 _ | 9 4 _ |
| _ 7 _ | _ 9 _ | _ _ 5 |
| 3 _ _ | _ _ 5 | _ 7 _ |
+-------+-------+-------+
| _ 8 7 | 4 _ _ | 1 _ _ |
| 4 6 3 | _ 8 _ | _ _ _ |
| _ _ _ | _ _ 7 | _ 8 _ |
+-------+-------+-------+
| 8 _ _ | 7 _ _ | _ _ _ |
| 7 _ _ | _ _ _ | _ 2 8 |
| _ 5 _ | 2 6 8 | _ _ _ |
+-------+-------+-------+

I can't confirm thats its solveable (yet) :)

cheers

Simon